The total dosage and therapeutic value of Cyclophosphamide (CPA) administered as adjuvant chemotherapy for breast cancer were studied. The subjects were those under 65 years of age who were diagnosed as Stage II breast cancer and underwent curative operation at any of 10 national hospitals from May, 1978 to April, 1981. CPA at dose of 100mg was administered orally every day. Comparison was made between the long-term administration group (L) of patients receiving a total of 25 g/m2 over 18 months and the short-term administration group (S) receiving a total of 6 g/m2 over 6 months, in terms of survival rate (SV) and disease-free survival rate (DFV). Cases selected as the subjects were randomly allotted to the two groups by means of the envelope method. The number of cases registered was 316. Of them, 308 cases were eligible for the study. Thirty-eight cases (23.6%) in L and 15 cases (10.2%) in S dropped out due to side effects and other reasons. No difference was observed between the two groups in background factors such as age, menopausal status, type of operation, histological diameter of tumor, histological type and metastasis to lymph nodes. Results of a 10-year follow-up revealed no difference in SV and DFV between the two groups. Excluding the drop-out cases, however, DFV was significantly high in L. There was no difference in DFV between the two group in terms of menopausal status. DFV was significantly high in L, if one includes cases with metastasis to lymph nodes. DFV was high in L, even when a group with no more than 3 metastatic lymph nodes and the group with not less than 4 were included. It is of little value to administer CPA to all Stage II cases for a long period of time. CPA may be indicated for cases showing metastasis to lymph nodes.